Abstract
Osteosarcoma is one of the most common malignant bone tumors mostly found in children and teenagers. Manual detection of osteosarcoma requires expertise and it is a labour-intensive process. If detected on time, the mortality rate can be reduced. With the advent of new technologies, automatic detection systems are used to analyse and classify medical images, which reduces the dependency on experts and leads to faster processing. In this paper, an automatic detection system: Integrated Features-Feature Selection Model for Classification (IF-FSM-C) to detect osteosarcoma from the high-resolution whole slide images (WSIs) is proposed. The novelty of the proposed approach is the use of integrated features obtained by fusion of features extracted using traditional handcrafted (HC) feature extraction techniques and deep learning models (DLMs) namely EfficientNet-B0 and Xception. To further improve the performance of the proposed system, feature selection (FS) is performed. Here, two binary variants of recently proposed Arithmetic Optimization Algorithm (AOA) known as BAOA-S and BAOA-V are proposed to perform FS. The selected features are given to a classifier that classifies the WSIs into Viable tumor (VT), Non-viable tumor (NVT) and non-tumor (NT). Experiments are performed to compare the performance of proposed IF-FSM-C to the classifiers which use HC or deep learning features alone as well as state-of-the-art methods for osteosarcoma detection. The best overall accuracy of 96.08% is obtained when integrated features extracted using HC techniques and Xception are used. The overall accuracy is enhanced to 99.54% after applying BAOA-S for FS. Further, the application of BAOA-S for FS reduces the number of features with the best model having only 188 features compared to 2118 features if no FS is applied.
Highlights
Osteogenic Sarcoma or Osteosarcoma is a form of bone cancer that typically starts growth in long bones of legs and arms [39]
We discuss the experimentation and analyse the results. 7.1 Experimental setup and results analysis Here, a publicly available dataset that contains Hematoxylin and Eosin (H&E) stained histological images of osteosarcoma [36] dataset comprises of 345 Viable Tumor (VT), 263 Non-Viable Tumor (NVT) and 536 NT images making a total of 1144 images is used
When feature selection (FS) is applied on features extracted using HC techniques and deep learning models (DLMs), it is observed from Tables 2 and 3 that it results in improvement in the performance of the classifier in case of both Binary Arithmetic Optimization Algorithm (BAOA)-S and BAOA-V
Summary
Osteogenic Sarcoma or Osteosarcoma is a form of bone cancer that typically starts growth in long bones of legs and arms [39]. In this case Convolutional Neural Networks (CNNs) are employed as the DLM for the task of feature extraction These models have the ability to automatically learn the features from the input image, but require a large training set with high variation for sufficient quality of attribute derivation. They have the ability to obtain higher number of low-level features which can describe the image in greater detail, helping in further analysis. Metaheuristic algorithms have been applied by researchers in many applications and obtained optimal solutions to the real-world problems which were difficult to solve or required a lot of processing time using conventional algorithms.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have